Obesity is associated with noninsulin-dependent diabetes mellitus (NIDDM) and coronary heart disease (CHD), and these interactions have usually been related to changes in immunoreactive insulin (IRI) levels. A role of proinsulin (PI) in this association has been suggested. We, therefore, examined IRI, PI, and true insulin levels and the PI/IRI ratio by glucose tolerance or CHD status in a cross-sectional study of 170 Japanese-American men (45-74 yr old) in whom 2 measures of adiposity (body mass index and intraabdominal fat) were made to assess potential associations in this population with a high prevalence of both NIDDM and CHD. Subjects were classified as having normal glucose tolerance (n = 58), impaired glucose tolerance (IGT; n = 55), or NIDDM (n = 57) or were classified by CHD status (without CHD, n = 127; with CHD, n = 43). A positive linear relationship existed between obesity, determined either as the body mass index or intraabdominal fat, and IRI, PI, and true insulin, but not the PI/IRI ratio. In the NIDDM subjects, PI levels were disproportionately greater than those in subjects with normal glucose tolerance or IGT, so the PI/IRI ratio was significantly greater in the NIDDM group [mean (95% confidence interval): normal glucose tolerance, 11.8% (range, 10.4-13.5); IGT, 12.8% (range, 10.8-15.1); NIDDM, 19.2% (range, 15.4-24.0); P = 0.0002] even when adjusted for obesity (P = 0.0004). In subjects with CHD compared to subjects without CHD, IRI (P = 0.0026) and true insulin levels (P = 0.0043) were increased, but PI levels were not. However, these differences were not present after adjustment for obesity. In contrast, when intraabdominal fat was adjusted for IRI or true insulin, a significant effect of intraabdominal fat on CHD risk was maintained (P = 0.045 and P = 0.029, respectively), suggesting that another factor(s) associated with central obesity may be involved in CHD risk. Thus, in Japanese-American men, elevated PI and PI/IRI ratio are markers of B-cell dysfunction, and these are not the result of obesity. An elevated true insulin level is present in those with CHD, but this appears to be the result of obesity. In contrast, central adiposity confers an additional risk for CHD independent of insulin.